Paper Title
Classifying Face Features for Better Recognition and Detection

Facial recognition underpins many real-time security systems. Face recognition (FR) models failed due to reduced resolution, occlusion, lighting, noise, and position fluctuation. Thus, by offering a unique AdaBoost Viola–Jones algorithm (AVJA) and Superficial Shallow Convolutional Neural Network (SSCNN) methodology, an effective FDR (face detection and recognition) system is developed. Method initially recognises image faces by calculating its global facial models in different locations and postures, then uses SSCNN to improve identification. Initially, the AVJA conducted facial detection (FD). The proposed AVJA handles unconstrained face images efficiently with boundedness, invariance, and image reconstruction. For FR, SSCNN is used to learn complex face-detected image features. Next, RA (recognition accuracy), and AP (average precision), AUC (area under curve) are compared to other approaches. The proposed approach detected facial images with 98.15% accuracy, outperforming existing methods. Keywords - Face Detection, Face Recognition, Face Features, Real-Time Applications.